cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
311
Views
0
Helpful
2
Replies

Inconsistent WAN performance between a 6506 and 3850

I have a point-to-point 100 Mbps circuit that I use to route traffic between a 6506 in Covina, CA and a 3850 stack in Raleigh, NC. The performance of VDI terminals in NC connecting to CA has been poor.  I used iperf to check the bandwidth for a single connection at various window sizes.

NC side is Windows 2008 Server
CA side is a CentOS 6.5

Winsize is TCP window size in KB -- both sides were set to the same value
RTT is round trip measured using ping from NC
Calulated value is (TCP window size)/RTT in Mbps
NC2CA is the measured bandwith sending from NC to CA in Mbps
CA2NC is the measured bandwith sending from CA to NC in Mbps

Winsize RTT(ms) Calculated  NC2CA  Ratio(%)  CA2NC Ratio(%)
64      68      7.35        1.18   16%       7.42  101%
128     68      14.71       1.15   8%        14.9  101%
256     68      29.41       1.17   4%        30.0  102%
512     68      58.82       1.22   2%        59.9  102%
1024    68      117.65      1.22   1%        118   100%
2048    68      235.29      1.45   1%        235   100%
4096    68      470.59      1.60   0%        467   99%
8192    68      941.18      2.25   0%        900   96%
16384   68      1882.35     3.22   0%        899   48%

The queuing strategy for the ports at either end of the point-to-point and to the respective servers is set to FIFO.

Time Warner insists their circuit is working correctly so I am trying to find something wrong on the 6506 and/or the 3850 switches.  Problem is, everything looks right to mr.  However, I am no expert so I hope someone here can tell where to start looking.

port on the 6506

#show interfaces gi1/3/9
GigabitEthernet1/3/9 is up, line protocol is up (connected)
  Hardware is C6k 1000Mb 802.3, address is 70ca.9b65.3bc0 (bia 70ca.9b65.3bc0)
  MTU 1500 bytes, BW 1000000 Kbit, DLY 10 usec,
     reliability 255/255, txload 1/255, rxload 1/255
  Encapsulation ARPA, loopback not set
  Keepalive set (10 sec)
  Full-duplex, 1000Mb/s, media type is 10/100/1000BaseT
  input flow-control is off, output flow-control is on
  Clock mode is auto
  ARP type: ARPA, ARP Timeout 04:00:00
  Last input 19w4d, output never, output hang never
  Last clearing of "show interface" counters 4d19h
  Input queue: 0/2000/0/0 (size/max/drops/flushes); Total output drops: 0
  Queueing strategy: fifo
  Output queue: 0/40 (size/max)
  5 minute input rate 719000 bits/sec, 567 packets/sec
  5 minute output rate 3524000 bits/sec, 632 packets/sec
     53678124 packets input, 9273723417 bytes, 0 no buffer
     Received 214987 broadcasts (214983 multicasts)
     0 runts, 0 giants, 0 throttles
     0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
     0 watchdog, 0 multicast, 0 pause input
     0 input packets with dribble condition detected
     81201059 packets output, 77569738141 bytes, 0 underruns
     0 output errors, 0 collisions, 0 interface resets
     0 babbles, 0 late collision, 0 deferred
     0 lost carrier, 0 no carrier, 0 PAUSE output
     0 output buffer failures, 0 output buffers swapped out

Port on the 3850

#show interface gi 1/0/1
GigabitEthernet1/0/1 is up, line protocol is up (connected)
  Hardware is Gigabit Ethernet, address is 189c.5dda.6b81 (bia 189c.5dda.6b81)
  MTU 1500 bytes, BW 1000000 Kbit/sec, DLY 10 usec,
     reliability 255/255, txload 1/255, rxload 1/255
  Encapsulation ARPA, loopback not set
  Keepalive set (10 sec)
  Full-duplex, 1000Mb/s, media type is 10/100/1000BaseTX
  input flow-control is off, output flow-control is unsupported
  ARP type: ARPA, ARP Timeout 04:00:00
  Last input 00:00:39, output never, output hang never
  Last clearing of "show interface" counters never
  Input queue: 0/2000/0/0 (size/max/drops/flushes); Total output drops: 0
  Queueing strategy: fifo
  Output queue: 0/40 (size/max)
  5 minute input rate 2931000 bits/sec, 545 packets/sec
  5 minute output rate 651000 bits/sec, 520 packets/sec
     158108261 packets input, 2247787225 bytes, 0 no buffer
     Received 18501 broadcasts (17419 multicasts)
     0 runts, 0 giants, 0 throttles
     0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
     0 watchdog, 17419 multicast, 0 pause input
     0 input packets with dribble condition detected
     114985635 packets output, 1843157338 bytes, 0 underruns
     0 output errors, 0 collisions, 1 interface resets
     0 unknown protocol drops
     0 babbles, 0 late collision, 0 deferred
     0 lost carrier, 0 no carrier, 0 pause output
     0 output buffer failures, 0 output buffers swapped out

 

2 Replies 2

rpinon
Level 1
Level 1

 I once had a similar problem,  ended up hardcoding the speed on the local ethernet interfaces, cleared the issue, somethign with clocking with the carrier. Don't ask me why,, worth a try?

Thanks.

That was the first thing T/W asked me to do.  I set the interfaces at both end to 1000/full with no success.

The only change so far is that TW did manage to throttle the 900 Mbps connection back to <100Mbps.

 

Getting Started

Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community:

Review Cisco Networking products for a $25 gift card